Local Path Planning among Pushable Objects based on Reinforcement Learning

TL;DR

This paper presents a reinforcement learning-based method for robot local path planning among pushable objects, capable of handling dynamic obstacles and real-world uncertainties through a trained policy tested in simulation and on a quadruped robot.

Contribution

It introduces a novel RL approach using Advantage Actor-Critic for obstacle pushing in confined spaces, adaptable to unseen scenarios and real-world conditions.

Findings

Effective obstacle pushing in simulated environments.

Successful deployment on a real quadruped robot.

Adaptability to dynamic and unpredictable obstacle behavior.

Abstract

In this paper, we introduce a method to deal with the problem of robot local path planning among pushable objects -- an open problem in robotics. In particular, we achieve that by training multiple agents simultaneously in a physics-based simulation environment, utilizing an Advantage Actor-Critic algorithm coupled with a deep neural network. The developed online policy enables these agents to push obstacles in ways that are not limited to axial alignments, adapt to unforeseen changes in obstacle dynamics instantaneously, and effectively tackle local path planning in confined areas. We tested the method in various simulated environments to prove the adaptation effectiveness to various unseen scenarios in unfamiliar settings. Moreover, we have successfully applied this policy on an actual quadruped robot, confirming its capability to handle the unpredictability and noise associated with real-world sensors and the inherent uncertainties present in unexplored object pushing tasks.